Louisiana State UniversityLSU Digital Commons
LSU Historical Dissertations and Theses Graduate School
1971
The Relationship Between Fixed Ratio Schedulesof Reinforcement and Aggression in Children.Gerald Leroy PetersonLouisiana State University and Agricultural & Mechanical College
Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses
This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion inLSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please [email protected].
Recommended CitationPeterson, Gerald Leroy, "The Relationship Between Fixed Ratio Schedules of Reinforcement and Aggression in Children." (1971).LSU Historical Dissertations and Theses. 1942.https://digitalcommons.lsu.edu/gradschool_disstheses/1942
71-20,614
PETERSON, Gerald LeRoy, 1944-THE RELATIONSHIP BETWEEN FIXED RATIO SCHEDULES OF REINFORCEMENT AND AGGRESSION IN CHILDREN.
The Louisiana State University and Agricultural and Mechanical College, Ph.D., 1971 Psychology, clinical
University Microfilms, A XEROX Com pany, Ann Arbor, Michigan ;
THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED
THE RELATIONSHIP BETWEEN FIXED RATIO SCHEDULES OF
REINFORCEMENT AND AGGRESSION IN CHILDREN
A Dissertation
Submitted to the Graduate Faculty of the Louisiana State University and
Agricultural and Mechanical College in partial fulfillment of the
requirements for the degree of Doctor of Philosophy
in
The Department of Psychology
byGerald L. Peterson
B.A., Seattle Pacific College, 1965 M.A., Louisiana State University, 1967
January, 1971
ACKNOWLEDGEMENTS
The author wishes to express his gratitude to all those whose
help and support have made this research possible.
He wishes to thank Dr. Edwin 0. Timmons for his continual encour
agement, guidance and valuable assistance that made this study possible.
To Dr. Bill Seay, Dr. Ralph Dreger, Dr. Laurence Siegel and Dr.
William Haag he expresses his sincere appreciation for their sugges
tions and guidance in the preparation of this study.
The writer is grateful to the directors, teachers, and
children of the University Methodist Nursery School in Baton Rouge,
Louisiana, whose cooperation and assistance made it possible for the
data to be collected.
He wishes to thank Miss Karen Calhoun for her continual support
and Mrs. Jewel Osbey for her typing of the original manuscript.
The author wishes to thank Mrs. Mary Mevers for her excellent
typing of the completed dissertation.
ii
TABLE OF CONTENTS
PAGE
TITLE P A G E ....................................................... i
ACKNOWLEDGEMENTS................................................... ii
LIST OF TABLES..................................................... iv
LIST OF FIGURES................................................... v
ABSTRACT........................................................... vi
INTRODUCTION ..................................................... 1
METHOD - EXPERIMENT I .......... 11
Subjects..........................................................11
Apparatus..........................................................11
Procedure..........................................................12
RESULTS - EXPERIMENT I ............................................. 14
METHOD - EXPERIMENT I I ............................................. 15
Subjects..........................................................15
Apparatus..........................................................15
Procedure..........................................................15
Analysis..........................................................15
RESULTS - EXPERIMENT II............................................. 17
DISCUSSION......................................................... 28
REFERENCES......................................................... 34
V I T A ................................................................39
iii
LIST OF TABLES
TABLE PAGE
1. T-Test of Amount of Time Spent in Aggressionduring FR50........................................... 25
2. T-Test of Amount of Time Spent in Aggressionduring Final Baseline................................. 27
iv
LIST OF FIGURES
FIGURE PAGE
1. Amount of Time Spent in Aggression for EightAggressing Subjects .................................... 18
2. Number of Pre-FR50 Lever Presses.......................... 21
3. Number of Pre-FR50 Reinforcements........................ 22
4. Average Amount of Time Spent Aggressing for Subjectswith Peak Aggression at FR50........................... 24
v
ABSTRACT
Two experiments using 24 pre-school children were performed in
an attempt to determine if results of aggression-inducing properties of
schedules of reinforcement on animals were generalizable to children.
Both experiments utilized a lever press task and a Bobo doll as the
target for aggression. Length of time spent aggressing was the depen
dent variable. In Experiment I where 8 males were reinforced 10 times
at levels CRF, FRIO, FR25, FR50, FR25, FRIO, and CRF only one S
aggressed and then only at the FR50 and FR25 levels. In Experiment II
schedules were built to FR50 as rapidly as possible comparing 8 boys
vs 8 girls. Eight of the 16 _Ss aggressed; the peak aggression occurred
at FR50. Although boys and girls were not significantly different at
the FR50 level the boys were significantly more aggressive during the
final baseline. It was concluded that high or "straining" FR schedules
are capable of inducing aggression equally in boys and girls. Aggression
appears to be a function of speed of transition to higher level
schedules as well as the response requirements of various high level
schedules of reinforcement.
vi
INTRODUCTION
Almost since Psychology's inception, aggression has been a focus
of attention. Why one human or animal behaves aggressively toward
another has been attributed to a number of things. Among these are:
Instinct (Lorenz, 1966);
Drive (Freud, 1921; Bender, 1953; Redl, 1957);
Child rearing (Sears, Whiting, Nowlis, and Sears, 1953);
Frustration (Dollard, Doob, Miller, Mowrer and Sears, 1939;
Miller, 1941; Dollard, 1944);
Instrumental responding (Skinner, 1959; Buss, 1961; Patterson,
Littman and Bricker, 1967);
Modeling (Bandura and Walters, 1963);
Aversive stimulation (Ulrich and Azrin, 1962; Azrin, Hutchinson
and Hake, 1963; Azrin, Ulrich, Hutchinson and Norman,
1964; Azrin, Hutchinson and Sallery, 1964; Azrin,
Hutchinson and McLaughlin, 1965; Azrin, Hake, and
Hutchinson, 1965; Boshka, Weisman and Thor, 1966);
Extinction (Gallup, 1965; Thompson and Bloom, 1966; Azrin,
Hutchinson and Hake, 1966; Davis and Donenfeld, 1967); and
Schedules of reinforcement (Hutchinson, Azrin and Hunt, 1968;
Gentry, 1968; Gentry and Schaeffer, 1969; Flory, 1969;
Knutson, 1970).
The concept of an instinct or drive as the basis for aggression
2
has been put forward on the one hand by Lorenz (1966) and, on the other,
by Freud and his psychoanalytic followers (Ross and Abrams, 1965).
Lorenz sees aggressive behavior as functional in the preservation of the
species. Although the aggressive instinct can be supressed to some
extent by ritualistic behavior, it still exists in a full blown state
in a particular animal or human and is ready to be set off by the appro
priate stimulus.
Freud (1921) and his followers (Ross and Abrams, 1965) also
postulate an aggressive instinct although it is decidedly different
from that proposed by Lorenz. In Freudian theory, while the sexual
instinct is of primary importance, aggression came to be regarded as the
second major instinct and his later writings discuss the life instinct
(Eros) and death instinct (Thanatos). In this dual instinct theory,
aggression served as an external manifestation of an internally directed
self destructive tendency (Thanatos). Aggression itself is referred to
in a variety of ways, raising the question of definition. Freud and his
followers discuss aggression in a number of different ways: as
achieving discharge in neurotic symptoms, as being self critical and
punitive in the super-ego component, as sublimation leading to control
over nature, and as fusing with the libido to produce significant
developmental and pathological phenomena (Freud, 1930; Ross and Abrams,
1965). Such a formulation of aggression makes it very difficult to
empirically test the concept.
Both Lorenz1 and Freud's concepts of aggression presuppose a
reservoir of instinctual aggressive energy seeking outlet. Berkowitz
3
(1967), in a number of studies on aggression, showed that such a con
ceptualization could not explain his data. He found that aggression
occurred only when certain stimulus events were present, and then only
under specifiable conditions.
A similar approach to that of Freud and Lorenz has been that
of physiologists who attempted to determine the physiological deter
minants of aggressive behavior. Beach (1945) and Beeman (1947) found a
direct relationship between hormonal balance and fighting behavior.
However, several studies since that time have downgraded the importance
of this variable. Bevan, Daves and Levy (1960) showed that testosterone
is less effective in producing fighting behavior in rats than is previous
fighting experience. Scott and Fredericson (1951) indicated that
hormonal factors are unimportant in fighting behavior following its
acquisition. Scott (1958) successfully trained mice not to fight and
concluded that the concept of a fighting instinct in mammals is unjust
ified. McNeil (1959) extended this conclusion to human beings, sug
gesting that, the explanation of human aggression is not furthered by
invoking genetic or physiological factors.
Thus, although early experiments, observations and speculations
suggested an instinctual or drive theory of aggression, these have not
been supported by later developments and alternative explanations have
been developed. Probably the most famous of these was the Frustration-
Aggression (F-A) hypothesis of Dollard, Doob, Miller, Mowrer and Sears
(1939) which developed from Freud's earlier writings. The F-A
hypothesis defined frustration as "that condition which exists when a
4
goal-response suffers interference" (p. 11) and aggression as "an act
whose goal response is injury to an organism (or organism surrogate)"
(p. 11). Their hypothesis suggested a one-to-one relationship between
frustration and aggression, that is "aggression is always a consequence
of frustration" (p. 1), and "the existence of frustration always leads
to some form of aggression" (p. 1). Later modifications of the
hypothesis regarded aggression as the naturally dominant but not neces
sarily inevitable consequence of frustration (Miller, 1941; Dollard,
1944). Thus nonaggressive behavior might occur if previous aggressive
responses had been unrewarded or punished. However, frustration con
tinued to be conceptualized as the inevitable antecedent of aggression.
Opposition to the hypothesis arose from numerous sources.
Bateson (1941) showed that cross^culturally this hypothesis was not
workable. In some cultures such as the Balinese, aggression was not
the typical response to frustration. Others (Barker, Dembo and Lewin,
1941; Wright, 1942, 1943) have shown that responses such as regression
may result from frustration. Likewise, results from other studies
(Maslow, 1941; Rosenweig, 1944; Buss, 1961) have shown that attack or
threat is more likely to result in aggression than is simply blocking
an ongoing response-sequence.
Although the F-A hypothesis may have limited applicability, it
cannot account for a great deal of aggression as was originally
intended. In an attempt to further the understanding of aggression,
new approaches have been formulated.
Bandura and Walters (1963) have reformulated the definitions of
5
aggression and frustration and by so doing, the F-A hypothesis. They
see aggression as "the class of pain-producing or damage-producing
responses, or as responses that could injure or damage if aimed at a
vulnerable object" (p. 366). Frustration was viewed as "a delay of
reinforcement" (p. 367). In their review of aggression, frustration is
seen as neither a necessary nor sufficient cause for aggressive behavior
to occur. The more important antecedents of aggression were seen as
past behavior with respect to prior reinforcement for aggression, and
exposure to aggressive models. The role of frustration appears to be
that it may elicit high magnitude responses which may be labeled aggres
sive if they are directed toward another person or animal. A study by
Walters and Brown (1964) supports this hypothesis. Under two separate
conditions, kindergarten through second grade boys were trained to hit
a Bobo doll with high or low intensity. A second group was trained on
a lever press which activated a ball in an enclosure. The boys in this
group were trained to activate the lever with high or low intensity
(i.e., under high intensity conditions the boys were rewarded for
pressing the lever hard enough for the ball to hit the top of the
enclosure). The subjects were then allowed to play games, likely to
elicit aggression, with another child. Both groups of boys trained under
high intensity were significantly more aggressive (as rated by frequency
of displayed aggressive behavior) than boys trained under low intensity
conditions. The authors point out that it was unlikely that frustra
tion in the training conditions accounted for this difference in
performance because although high intensity children trained on the Bobo
6
doll were reinforced for only 1 out of 12 responses and low intensity
boys were reinforced for 1 out of 8 responses, the reverse was true for
the lever press group.
Support for the contention that modeling and previous history of
reinforcement are more important variables in the display of aggressive
behavior than is frustration have come from a number of sources. Bandura
and Huston (1961) showed that preschool children exposed to a model who
performed aggressive acts displayed significantly more aggressive
behavior than did a control group exposed to the same model but without
the demonstration of aggressive behavior. Half of the experimental
group had experienced two periods of rewarding interaction with the
model while the other half spent two periods of time with the model but
without interacting with her. Although the rewarded group imitated
more of the model's behavior, both groups showed a high incidence of
aggression in relation to the control group, suggesting that merely
observing an aggressive model is sufficient to produce aggressive
behavior in children. Bandura, Ross and Ross (1961) extended this
finding by showing that aggressive imitative responses developed in the
presence of a model generalize to settings in which the model is absent.
In another study (Bandura, Ross and Ross, 1962) they showed that film-
mediated aggressive modeling led to aggressive behavior in the observ
ing children.
In a somewhat related study by Schacter and Singer (1962),
subjects injected with epinephrine and not informed of its physiologi
cal side effects displayed considerably more aggressive behavior toward
7
an aggressive confederate than a group similarly injected and informed
of the side effects. Bandura and Walters (1963) interpret these
results as suggesting that the influence of models is more potent under
emotional arousal especially when the subjects cannot attribute their
feelings to anything other than a model's behavior.
Bandura and Walters (1963) conclude that observation of modeling
behavior has two different effects: 1) the observer may acquire new
responses that did not previously exist in his behavior repertory, or
2) observation of aggressive models may weaken inhibitory responses.
Under this second condition behavior which the model exhibits already
exists in his behavior repertory and may be nonimitative in nature.
Reinforcement of aggressive behavior is seen by Bandura and
Walters (1963) as the second important variable in current aggressive
behavior. A number of studies have demonstrated its importance. Davis
(1943) and Davis and Havighurst (1947) reported that lower class parents
encourage and reward aggression to a greater extent than do middle
class parents, thus perhaps accounting for the higher incidence of
aggressive behavior in the lower classes. Bandura and Walters (1959;
1963) discovered that parents of aggressive boys encouraged and condoned
aggressive behavior more than parents of non-aggressive boys.
In addition to parental importance in shaping aggressive
behavior, Patterson, jet al. (1967) have shown the influence of peer
behavior. In one of the few aggression studies with children done out
side the laboratory, Patterson sent into two nursery school settings a
corps of observers who transcribed behavioral events and their
8
consequences. In later analyses of these data he was able to show that
much of the aggressive behavior was being reinforced by peers at a very
high level and was thus being maintained. Based on his results a sub
sequent experiment was performed showing that aggressive behavior could
be predicted quite accurately if knowledge of the immediately previous
aggressive behavior and its consequences were known.
Several other studies have been done showing the effects of
reinforcement of aggressive responses. Cowan and Walters (1963) found
that reinforcing boys for aggressive behavior on three different
schedules of reinforcement (CRF, FR3, FR6) resulted in typical extinc
tion curves, that is, CRF boys extinguished the fastest and the FR6
boys took the longest time to extinguish. In another study, Walters
and Brown (1963) used four conditions: an FR6 and a CRF schedule of
reinforcement for hitting a Bobo doll and a no training group. The
results showed that the FR6 group exhibited the greatest amount of
aggressive behavior and was significantly different from the other
three groups who did not differ significantly from one another. An
additional variable, frustration-nonfrustration, made no difference in
the amount of aggression exhibited.
Although it appears that a great proportion of aggressive
behavior can be accounted for by the influence of models and operant
conditioning, other studies have shown the importance of aversive
stimulation in eliciting aggression. This phenomenon was first
described by O'Kelly and Steckle (1939) and has more recently been
confirmed and expanded (Ulrich and Azrin, 1962). Aversive stimulation
9
has been shown to elicit aggressive behavior from a variety of species
(Ulrich & Azrin, 1962; Azrin, et al., 1963); Ulrich, Hutchinson and
Azrin, 1965). The aversive stimulus may be of an exteroceptive nature
such as foot shock, heat (Ulrich and Azrin, 1962) or tail pinch
(Azrin, ef aJL., 1965), or it may be interoceptive such as morphine
withdrawal (Boshka, ejt al., 1966).
Ulrich and Azrin (1962) conclude from their studies that the
aggression exhibited was not of an operant nature (shock resulted in
attack rather than leaning or climbing on other animals to eliminate
the shock), nor was it superstitious, as continuous delivery of shock
produced attack behavior. Azrin, e£ al., (1964) regard the pain-
aggression reaction as a type of reflexive reaction and not as a
response maintained by operant reinforcement.
More recent investigations have suggested that extinction and
non-reinforced trials are sufficient to result in aggressive behavior
with paired organisms. Rats will attack one another in a straight alley
on non-reinforced trials (Gallup, 1965). Thompson and Bloom (1966)
and Davis and Donenfeld (1967) report that rats placed on extinction
following a CRF schedule of reinforcement will aggress. Azrin, et al.,
(1966) showed that aggressive behavior during extinction was not
specific to rats but extended to pigeons as well.
Since one of the characteristics of intermittent schedules of
reinforcement are periods of non-reinforcement, Azrin, et al,, (1966)
suggested that intermittent schedules of reinforcement might elicit
aggressive behavior between pairs of subjects. This hypothesis has been
10
supported by a number of studies. Gentry (1968), Flory (1969) and
Knutson (1970) found marked aggression in pigeons using relatively high
(40-120) FR schedules of reinforcement. Gentry and Schaeffer (1969)
found this effect to be present in rats, while Hutchinson, cjt a_l., (1968)
found that squirrel monkeys exhibited aggressive responses toward a
pneumatic hose when subjected to high requirements of FR responding.
The purpose of the present study was to determine whether
aggressive behavior resulting from high response requirement FR schedules
of reinforcement are generalizable to human subjects. This was accom
plished by using pre-school children in an experimental situation
requiring varying numbers of lever press responses and utilizing a Bobo
doll as a target for aggression. The findings of this study should
result in added information to our knowledge of aggressive behavior.
The results should also be helpful in the study of children and the
treatment procedures used with them. If, for example, a therapist is
seeing an aggressive child he might look at the schedules of reinforce
ment the child is receiving in addition to other antecedent conditions
known to elicit aggression. For behavior modifiers, especially, the
understanding of the interaction between schedules of reinforcement and
the presence of stimuli in the environment is essential for adequate
treatment.
METHOD
Experiment I
Subjects
The S,s were 8 five year old nursery school boys.
Apparatus
The experimental chamber was a screened-off portion of a larger
room. The chamber contained a small chair, a small table, the stimulus-
reinforcement box and a Bobo doll. The doll was contained in a three
sided, box-like enclosure. The control room also located in the larger
room contained the stimulus-reinforcement box control panel and an
Esterline Angus 20 channel event recorder connected with the stimulus-
reinforcement box and the Bobo doll enclosure.
The stimulus reinforcement box was a black box 19" x 11" x 15"
with a round stimulus light on the front panel. Directly below the
stimulus light was the manipulandum, a key which must be pressed by j3s.
Four inches to the left of the key was a reinforcement chute and a
plexiglas reinforcement catch box.
The Bobo doll enclosure was a 24" x 30" x 24" three sided box
with the front side open. The doll was placed in the front part of the
box, facing outward. Immediately behind the doll, on the floor of the
box was a 6" restraining wall which kept the doll from moving around the
enclosure excessively.
12
The control panel of the stimulus-reinforcement box contained a
switch for the stimulus light, a switch which determined the schedule of
reinforcement and a reinforcement reset button that controlled the dis
pensing of the reinforcement. The event recorder, which recorded the
number of lever presses, reinforcements and amount of time spent
aggressing was also placed in the control room.
Procedure
Initial pilot work indicated that displacements of the doll's
head would be an appropriate measure of aggression. Later pilot work
showed however, that this index was not sufficient, for other aggressive
responses such as wrestling and pinching were not automatically recorded
by the apparatus. In order to take these additional behaviors into
account it was decided to use time spent in aggression as the dependent
variable rather than displacements of the doll's head. E observed S[
through the screen and made a mark on the event recorder paper when E
judged aggression had begun and again when aggression terminated.
Responses labeled "aggressive" were based on the categories reported by
Walters and Brown (1964) and included, elbowing, kicking, punching and
pushing. In addition to these wrestling and squeezing or pinching were
included.
Each j3 was placed in the experimental chamber in front of the
stimulus-reinforcement box with the Bobo doll on the side of his pre
ferred hand for 3 minutes with the stimulus light darkened and the
response key inoperative. During this time the event^ recorder was active
and a baseline rate of hitting responses was recorded.
13
Following the time allowed for baseline recording, j3 was taken
from the experimental chamber and shown a number of items such as,
coloring books, small toy cars and trucks, kites, play dough, coupons
which could be redeemed at a local ice cream store, etc., and told that
if he worked real hard he could win one of the items. j3 was then
reintroduced to the experimental chamber and told that he would earn
some marbles which he could trade for the toys he had already seen.
S! was seated in front of the stimulus reinforcement box with the Bobo
doll on the side of his preferred hand. E demonstrated how j3 could earn
the marbles by pressing the key below the stimulus light and receiving
a marble. E then left the chamber and returned to the control panel.
was told to begin and was successively programmed through CRF, FRIO,
FR25, FR50, FR25, FRIO, CRF and a 3 minute baseline period. Each
schedule was continued until S_ received 10 reinforcements, at which
time the schedule was increased to the next FR requirement. The amount
of time spent in aggression was recorded for each schedule. During the
final baseline period the stimulus light was again darkened and the
response key inoperative. The event recorder remained operative.
RESULTS
Experiment I
The various levels of FR response requirements in this experi
mental situation failed to elicit aggression, as previously defined,
in all children but one. The one child who aggressed did so as
predicted; that is, he exhibited aggressive behavior at the higher FR
schedules (FR50, FR25) rather than during the lower ones. In addition,
one other child left the experimental chamber during FR50 but returned
and completed the task. No other children displayed aggressive
behavior at any time during the experimental sessions.
Two _Ss could not be used in the data analysis and were replaced
because they failed to respond beyond the FRIO level. This failure
made it impossible to build the higher schedules necessary for this
study. These two _Ss did, however, display aggressive behavior at the
FRIO level following no aggression during either the baseline or CRF
segments.
METHOD
Experiment II
Subjects
The Ss were 8 five year old nursery school boys and 8 five year
old nursery school girls.
Apparatus
The apparatus was the same as that described in Experiment I.
Procedure
The baseline and demonstration of equipment was identical to
that described in Experiment I. Each _S was then progressively shaped
to respond to an FR50 schedule of reinforcement. Typically a Si was
reinforced two or three times during CRF, FR2, FR3, FR4, FR5, FRIO, and
FR25 schedules. The schedule was increased only when a consistent rate
of responding occurred. Although practically no aggression was elicited
in Experiment I, previous animal studies (Gentry, 1968; Knutson, 1970)
have shown that FR50 schedules of reinforcement are sufficiently high to
induce aggression. Each received 20 reinforcements on this schedule
followed by a 3 minute baseline during which the stimulus light was
darkened and the response key inoperative. During this time the event
recorder remained in operation.
AnalysisThe data were analyzed with a t-test, as suggested by Hayes (1963),
16
to test differences between the two groups. The dependent variable to
be analyzed was the amount of time spent aggressing during the schedule
under study. A .05 level of significance was accepted as showing an
effect.
RESULTS
Experiment II
Aggression was elicited from eight of 16 .Ss in Experiment II.
Among the eight were wide individual differences in both amount of
aggression displayed and the point at which peak aggression occurred.
These differences can be seen in Figure 1 where individual performances
are presented.
In the pre-treatment baseline period the boys were slightly more
aggressive than the girls. Two boys aggressed for a total of 26 seconds
while no girls aggressed during this time.
The schedule building portion of this experiment also induced
only minimal aggression. No boys and only one girl aggressed during
this period. The j3s varied widely in the number of responses necessary
to reach a steady rate of responding. The range of responses are found
in Figure 2. Responses ranged from 95 to 191 with a mean for the girls
of 131.5 and for the boys 127.3. The range of reinforcements received
for each _S are shown in Figure 3. The total number of reinforcements
received varied from 13 to 23, with the girl's mean 17.5 and the boy's
mean 18.5. As can be seen j3s who aggressed ranged from the least
number of lever presses before the FR50 segment to the greatest number
of responses. They also ranged from the least to the greatest number
of reinforcements.
Seco
nds
18
60
50
40
30
20
10
0Base Build FR Baseline 50 Line
60
50
40
30
20
Build FRBase Baseline 50 line
Segments
Amount of Time Spent in Aggression for Eight Aggressing Subjects
Figure 1.
Seco
nds
19
20
Base Build FR Base
30 B1
20
10
Base Build FR Baseline 50 line line line
Segments
60
50
40
30
20
10
Base Build FR Base
G460
50
40
30
20
10
Base Build FR Baseline 50 line line 50 line
Segments
Amount of Time Spent in Aggression for Eight Aggressing Subjects
Figure 1 (Continued)
130
120
110
100
90
80
70
60
50
40
30
20
10
0 -
20
130 G8
120
110
100
90
80
60
50
40
30
20
10
Base Build ire! Base line 50 line
Base Build FR Base line 50 line
Segments
Amount of Time Spent in Aggression for Eight Aggressing SubjectsFigure 1 (Continued)
21
Hi Agsre □
ssors
Nonaggressors
M100- 110- 120- 130- 140- 150- 160- 170- 180- 190-109 119 129 139 149 159 169 179 189 199
lever presses
Number of pre-FR50 Lever Presses
Figure 2
Subj
ects
22
issorsAggrej
| j Nonaggressors
I13 14 15 16 17 18 19 20 21 22 23
reinforcements
Number of pre-FR50
Reinforcements
Figure 3
23
During the FR50 segment of the experiment four boys and three
girls aggressed. One of the four boys hit the stimulus-reinforcement
box rather than the Bobo doll but, because of the similarity of
responses, was included in the aggressive group. For six of the seven
children peak aggression was displayed during this segment. Figure 4
shows the average amount of time spent in aggression by this group for
each segment of the experiment. The length of time spent in aggres
sion during the FR50 segment was analyzed twice and the results are
presented in Table 1. In the first comparison all _Ss were used. A
t-test was performed and no significant differences between boys and
girls were found. A second t-test was performed on the amount of
time spent in aggression of only the _Ss who aggressed, and again no
significant differences were found between boys and girls.
Subjects who aggressed during the FR50 segment varied as to the
point at which aggression began. Two Ss commenced aggressing before
the first reinforcement was dispensed, one JS began aggressing between
the second and third reinforcement, two j3s began between the third and
fourth reinforcement, one began following the fourth reinforcement and
one £> waited until 12 reinforcements had been presented before
aggressing. The location of the aggression was largely during the S/s
response run. One j> aggressed only during the post-reinforcement-
pause and one j> combined aggression during the post-reinforcement-
pause with aggression during the response run. All other Sis who
aggressed did so during the response run itself.
Boys were found to be significantly more aggressive than girls
Second
s
24
60
50
40
30
20
10
0
baseline building FR50 baseline
Segments
Average Amount of Time Spent Aggressing for
Subjects with Peak Aggression at FR50
Figure 4
25
TABLE 1
T-TEST OF AMOUNT OF TIME SPENT IN
AGGRESSION DURING FR50
Group N MeanStandardDeviation t
All Subjects
Girls 8 22.3 41.8.51 (NS)
Boys 8 19.5 33.9
Aggressors
Girls 3 59.3 49.62.23 (NS)
Boys 4 39.0 42.3
26
during the post-baseline period. This difference is shown in Table 2
where the results of the t-tests between all Jls and between aggressors
only are shown.
Although only four boys exhibited aggression during the FR50
segment of the experiment, two others indicated that they wanted to hit
the doll during the FR50 segment but failed to do so because of their
desire to earn enough marbles to obtain a toy.
One female S! never reached a level rate of responding beyond
FR5 and was replaced for the data analysis.
In summary, aggression was elicited from eight of 16 J3s during
this experiment. Six J5s (three boys and three girls) showed peak
aggression during the FR50 schedule. The remaining two male j>s
aggressed most during the final baseline segment. Most of the aggres
sion commenced early in the FR50 schedule and usually occurred during
the response run.
27
TABLE 2
T-TEST OF AMOUNT OF TIME SPENT IN
AGGRESSION DURING FINAL BASELINE
Group N Mean StandardDeviation t
All Subjects
Girls 8 1.8 3.44.30*
Boys 8 14.9 23.0
Aggressors
Girls 2 7.0 3.03.97*
Boys 3 36.3 17.7
* p .05
DISCUSSION
It was shown that high FR schedules of reinforcement are capable
of inducing aggression in children. The extent of the generalization
at this point is still unknown as a human surrogate was used as a
target for aggression. The results are, however, consistent with prev
ious animal studies and suggest that their findings may have wider
implications than the results of the present study warrant.
Experiment II showed that relatively high FR schedules of
reinforcement will elicit aggression toward human surrogates from some
Ss. No consistent differences were found between those Sis who
aggressed and those who did not. Although individual £s varied a
great deal as to the number of pre-FR50 lever presses and reinforce
ments, aggressive j3s varied from the least number lever presses and
reinforcements to the greatest number of each and it is unlikely that
this variable accounted for the differences in elicited aggression.
The fact that only eight of 16 Sis aggressed is in agreement
with previous animal studies. Gentry (1968) found that with live
pigeon targets, only two of three experimental birds aggressed.
Gentry and Schaeffer (1969) found that one of four pairs of rats
failed to display aggression under different schedules of reinforce
ment. Knutson (1970) found that only one of five of his experimental
birds would attack a stuffed pigeon. Pain-induced aggression also
failed to produce aggressive behavior toward inanimate objects for all
29
rats (Ulrich and Azrin, 1962) or for all squirrel monkeys (Azrin et al.,
1964; Azrin, et^aJL., 1965).
Sex differences also failed to discriminate aggressors from non
aggressors. Previous studies (Bandura et al., 1961; 1963(a); 1963(b))
have shown that boys display significantly more aggression than girls.
The present study only partially supports these findings. Although
boys were slightly more aggressive during the initial baseline period,
there were no significant differences between boys' and girls' aggres
sive behavior during the FR50 schedule of reinforcement. During the
post-reinforcement baseline period however, the boys were significantly
more aggressive than the girls. This suggests that high FR schedules
of reinforcement are capable of eliciting aggression without respect
to sex. When this condition is removed however, boys continue to dis
play somewhat more aggression.
Although previous animal studies of the aggression inducing
properties of schedules of reinforcement have dealt only with male jSs,
the initial study (Ulrich and Azrin, 1962) of the effects of aversive
stimulation on aggression showed no sex differences in amount of
aggression displayed. It has been shown that high response requirement
schedules of reinforcement can be considered aversive (Azrin, 1961).
Thus it is not surprising that no sex differences were found in the
present study.
Two different patterns of aggressive behavior, peak aggression
during FR50 and peak aggression during the post-reinforcement baseline
period were found in the present study. This finding is also consistent
30
with the animal literature. The greatest amount of aggression (six _Ss)
occurred during the FR50 segment of the experiment and supports the
results of Gentry (1968), Hutchinson, et al, (1968), Flory (1969) and
Knutson (1970).
Azrin, et al. (1966) have shown that extinction following CRF
will result in aggressive behavior. Hutchinson, et al, (1968) suggested
that an intermittent reinforcement history might result in greater
attack during extinction than a CRF history. Knutson (1970) found only
one that showed increased aggression during extinction with in
creased FR response requirements. One £1 showed no change in extinc
tion elicited aggression and three jSs displayed less extinction-elicited
aggression with increased FR response requirements. The present study's
post reinforcement baseline period was very similar to the extinction
procedure described by Knutson (1970) and similar results were obtained.
Two Ss showed increased post reinforcement baseline (extinction) aggres
sion while six j>s showed reduced post reinforcement baseline (extinc
tion) aggression.
Experiment I, with the exception of one jj, failed to elicit
aggression, while eight of 16 j>s in Experiment II aggressed. The
different conditions involved in these experiments may account for this
difference. One difference was that the FR50 segment of Experiment II
was lengthened to 1000 responses and 20 reinforcements. The data show
that only one j> of seven who aggressed during the FR50 schedule
commenced aggression following the twelfth reinforcement. The remain
ing six Sis began aggressing well before the tenth reinforcement. A
31
second major difference was the way in which the FR50 segment was
reached. In Experiment I schedules were gradually raised from CRF to
FRIO, FR25, and FR50. The schedules were then gradually returned to
CRF in reverse order. In Experiment II the schedules were build as
rapidly as possible. This difference resulted in a differential of both
pre-FR50 responses and reinforcements. In Experiment I each S!
responded 360 times and received 30 reinforcements before reaching the
FR50 schedule. In Experiment II the pre-FR50 responses ranged from a
low of 95 to a high of 191, while the reinforcements varied from 13 to
23. Thus in terms of both responses and reinforcements, jSs in Experi
ment II went through a much more rapid transition than those in Experi
ment I.
Ferster and Skinner (1957) have shown that rapid transition from
CRF to higher FR levels result in strained responding at the higher
levels with individual _Ss. It is possible that when even surrogate
humans are in an individual's environment, aggression rather than simply
strained performance is the result of rapid transition from a lower FR
requirement to a higher one and the subsequent response requirements at
the higher FR schedule. This is supported by Knutson's (1970) findings
that the introduction of a target bird into the experimental chamber
resulted in extremely strained performance by the experimental pigeon in
addition to the time spent aggressing. These findings suggest that much
more attention must be given to the importance of the interaction
between schedules of reinforcement and the environment affecting a
particular _S.
32
The present study shows that aggressive behavior may occur in
young children as the result of high response requirements on an FR
schedule of reinforcement and thus may account for some of the aggres
sive actions observable in human behavior. Although Bandura and
Walters (1963) and Patterson (1967) have shown that a great deal of
aggression in children can be accounted for by the effects of modeling
and peer reinforcement, it is probable that a certain amount remains
unaccounted for. How much aggression can be accounted for by straining
schedules of reinforcement account for is still unknown; however, this
study has shown that aggression may very well be the result of high
response requirement schedules of reinforcement and further study is
warranted.
The present study only attempted to determine if the results of
animal studies of schedule of reinforcement-induced aggression were
generalizable to children. The results obtained are consistent with
the findings of previous studies in this area. Further study is now
necessary to determine what the parameters of schedule of reinforcement-
induced aggression are. Studies comparing different groups of people
would appear in order. Also more extended work should be done with
respect to the schedules themselves. Higher ratio schedules could be
used as well as temporal types of schedules. In addition the FR
schedules should be run over a longer period of time using numerous con
ditioning sessions.
Ultimately, studies should be done in a more naturalistic setting.
A particular child could be singled out and one aspect of his behavior
33
reinforced at a rate equivalent to FR50 or higher and the amount of
aggression exhibited toward other children recorded. In this way the
practical aspects of this knowledge can be more fully evaluated.
REFERENCES
Azrin,
Azrin,
Azrin,
Azrin,
Azrin,
Azrin,
Azrin,
Bandura
Bandura
Bandura
Bandura
f. H. Time-out from positive reinforcement. Science, 1961,133, 382-383.
I. H., Hake, D. F., & Hutchinson, R. R. Elicitation of aggression by a physical blow. Journal of the Experimental Analysis of Behavior, 1965, 8, 55-57.
I. H., Hutchinson, R. R. & Hake, D. F. Pain-induced fighting in the squirrel monkey. Journal of the Experimental Analysis of Behavior, 1963, 6, 620-621.
L H., Hutchinson, R. R., & Hake, D. F. Extinction-induced aggression. Journal of the Experimental Analysis of Behavior. 1966, 9, 191-204.
I. H., Hutchinson, R. R. & McLaughlin, R. The opportunity for aggression as an operant reinforcer during aversive stimula- tion. Journal of the Experimental Analysis of Behavior, 1965,8, 171-180.
f. H., Hutchinson, R. R., & Sallery, R. D. Pain-aggression toward inanimate objects. Journal of the Experimental Analysis of Behavior, 1964, 7, 223-227.
[. H., Ulrich, R. E., Hutchinson, R. R., & Norman, D. G. Effect of shock duration on shock-induced fighting. Journal of the Experimental Analysis of Behavior. 1964, 7, 9-11.
A., & Huston, A. Identification as a process of incidental learning. Journal of Abnormal and Social Psychology, 1961, 63, 311-318.
A., Ross, D., & Ross, S. A. Transmission of aggression through imitation of aggressive models. Journal of Abnormal and Social Psychology. 1961, 63, 575-582.
A., Ross, D., & Ross, S. A. Imitation of film-mediated aggressive models. Journal of Abnormal and Social Psychology, 1963, 66, 3-11 (a).
A., Ross, D., & Ross, S. A. Vicarious reinforcement and imitative learning. Journal of Abnormal and Social Psychology. 1963, 67, 601-607 (b).
35
Bandura, A., & Walters, R. H. Adolescent aggression. New York: RonaldPress, 1959.
Bandura, A., & Walters, R. H. Aggression. Yearbook of the National Society for the Study of Education. 1963, 62, 364-415.
Barker, R. G., Dembo, T., & Lewin, K. Frustration and regression: Anexperiment with young children. University of Iowa Studies on Child Welfare, 1941, 28, 1-314.
Bateson, G. The frustration-aggression hypothesis and culture. Psychological Review, 1941, 48, 350-355.
Beach, F. A. Bisexual mating behavior in the male rat: Effects ofcastration and hormone administration. Physiological Zoology, 1945, 18, 390-402.
Beeman, E. A. The effect of male hormone on aggressive behavior in mice. Physiological Zoology, 1947, 20, 373-405.
Bender, L. Aggression, hostility and anxiety in children. Springfield, 111.: Thomas, 1953.
Berkowitz, L. Experiments on automatism and intention in human aggression. In C. D. Clemente & D. B. Lindsley (Ed.), Aggression and defense, Berkeley: University of California Press, 1967.
Bevan, W., Daves, W. F., & Levy, G. W. The relation of castration,androgen therapy and pre-test fighting experience to competitive aggression in male C57BL/10 mice. Animal Behavior, 1960,8, 6-12.
Boshka, S. C., Weisman, H. M., & Thor, D. H. A technique for inducing aggression in rats utilizing morphine withdrawal. The Psychological Record, 1966, 16, 541-543.
Buss, A. H. The psychology of aggression. New York: Wiley, 1961.
Cowan, P. A., & Walters, R. H. Studies of reinforcement of aggression: Effects of scheduling. Child development, 1963, 34, 543-551.
Davis, A., Child training and social class. In R. G. Barker, J. Kounin, & H. F. Wright (Ed.), Child behavior and development. New York: McGraw-Hill, 1943.
Davis, A., & Havighurst, R. J. The father of the man. Boston;Hought on-Mi f flin, 1947.
Davis, H., & Donenfeld, I. Extinction induced social interaction in rats. Psychonomic Science, 1967, 7, 85-86.
36
Dollard, J. Frustration and aggression. London: Kegan Paul, 1944.
Dollard, J., Doob, L. W., Miller, N. E., Mowrer, 0. H. & Sears, R. R.Frustration and aggression. New Haven: Yale University Press,1939.
Ferster, C. B., 6c Skinner, B. F. Schedules of reinforcement. New York: Appleton-Century-Crofts, 1957.
Flory, R. K. Attack behavior in a multiple fixed-ratio schedule of reinforcement. Psychonomic Science. 1969, 16(3), 156-157.
Freud, S. Group psychology and the analysis of the ego, (1921). In The standard edition of the complete psychological works of Sigmund Freud. Vol. 18, London: Hogarth Press and the Institute of Psycho-Analysis, 1962.
Freud, S. Civilization and its discontents, (1930). In The standard edition of the complete psychological works of Sigmund Freud. Vol. 21, London: Hogarth Press and the Institute of Psycho-Analysis, 1962.
Gallup, G. G. Jr. Aggression in rats as a function of frustrativenonreward in a straight alley. Psychonomic Science. 1965, 3, 99-100.
Gentry, W. D. Fixed-ratiorschedule-induced aggression. Journal of the Experimental Analysis of Behavior. 1968, 11, 813-817.
Gentry, W. D., 6c Schaeffer, R. W. The effect of FR response requirement on aggressive behavior in rats. Psychonomic Science,1969, 14(5), 236, 238.
Hayes, W. L. Statistics for psychologists. New York: Holt, Rhinehart6c Winston, 1963.
Hutchinson, R. R., Azrin, N. H., 6c Hunt, G. M. Attack produced byintermittent reinforcement of a concurrent operant response. Journal of the Experimental Analysis of Behavior, 1968, 11, 489-495.
Knutson, J. F. Aggression during the fixed-ration and extinction components of a multiple schedule of reinforcement. Journal of the Experimental Analysis of Behavior, 1970, 13, 221-231.
Lorenz, K. On Aggression. New York: Harcourt, Brace 6c World, 1966.
Maslow, A. H. Deprivation, threat and frustration. PsychologicalReview. 1941, 48, 364-366.
37
McNeil, E. B. Psychology and aggression. Conflict Resolution. 1959,3, 195-293.
Miller, N. E. The frustration-aggression hypothesis. Psychological Review, 1941, 48, 337-342.
O'Kelly, L. E., & Steckle, L. C. A note on long-enduring emotionalresponses in the rat. Journal of Psychology. 1939, 8, 125-131.
Patterson, G., Littman, R. A., & Bricker, W. Assertive behavior inchildren; a step toward a theory of aggression. Monograph of the Society for Research in Child Development, 1967, 32(5).
Redl, F., & Wineman, D. The aggressive child. Glencoe: Free Press,1957.
Rosenzweig, S. An outline of frustration theory. In J. McV. Hunt (Ed.) Personality and the behavior disorders. New York:Ronald Press, 1944.
Ross, N., 6c Abrams, S. Fundamentals of psychoanalytic theory. In. B. Wolman (Ed.) Handbook of clinical psychology. New York: McGraw-Hill, 1965.
Schacter, S., 6c Singer, J. E. Cognitive, social and physiologicaldeterminants of emotional state. Psychological Review. 1962,69, 379-399.
Scott, J. P. Aggression. Chicago: University of Chicago Press, 1958.
Scott, J. P., 6c Fredericson, E. The causes of fighting in mice and rats. Physiological Zoology, 1951, 24, 273-309.
Sears, R., Whiting, J., Nowlis, V., 6c Sears, P. Some childrearing antecedents of aggression and dependency in young children. General Psychology Monographs. 1953, 47, 135-204.
Skinner, B. F. An experimental analysis of certain emotions. Journal of the Experimental Analysis of Behavior, 1959, 2, 264.
Thompson, T., 6c Bloom, W . ; Aggressive behavior and extinctioninducedresponse-rate increase. Psychonomic Science. 1966, 5, 335-336.
Ulrich, R. E., 6c Azrin, N. H. Reflexive fighting in response toaversive stimulation. Journal of the Experimental Analysis of Behavior. 1962, 5, 511-520.
Ulrich, R. E., Hutchinson, R. R. 6e Azrin, N. H. Pain-elicited aggres- sion. The Psychological Record. 1965, 15, 111-126.
38
Walters
Walters
Wright
Wright,
, R. H., & Brown, M. Studies of reinforcement of aggression:III transfer of responses to an interpersonal situation.Child Development, 1963, 34, 563-571.
, R. H., & Brown, M. A test of the high-magnitude theory of aggression. Journal of Experimental Child Psychology. 1964, 1, 376-387.
, M. E. Constructiveness of play as affected by group organization and frustration. Character and Personality. 1942, 11, 40-49.
M. E. The influence of frustration upon the social relations of young children. Character and Personality, 1943, 12, 111-112.
VITA
Gerald L. Peterson was born in Eagle Bend, Minnesota, on
January 24, 1944. He attended Bismarck High School through his junior
year and graduated from Roosevelt High School in Seattle in 1961.
Following graduation he entered Seattle Pacific College, Seattle,
Washington. He received the degree of Bachelor of Arts from that
institution in 1965 with a major in psychology. He entered Louisiana
State University in September 1965 and received his M.A. from that
institution in January, 1967. He held an assistantship in the depart
ment for the academic year 1965-66. In 1966-67 he was supported by
a United States Public Health Service Fellowship. For the academic
year 1967-68 he was a teaching assistant. He spent the following
year on internship at the University of Oregon Medical School. At
the present time he is the recipient of a Veteran's Administration
traineeship.
EXAMINATION AND THESIS REPORT
Candidate: Gerald L. Peterson
Major Field: Psychology
Title of Thesis: The Relationship between Fixed-ratio Schedules of Reinforcement and Aggression in Children
Approved:
O QMajor Professor and Chairman
Dean or the Graduate School
EXAMINING CO! ITTEE:
Date of Examination:
July 28j 1Q70